1. Field of the Invention
The present invention relates to a method for driving an image pickup element, an image pickup system utilizing such an element, and a recording medium storing a program for implementing such a method.
2. Description of Related Art
Japanese Patents Laid-Open Nos. 09-214836 and 2000-032318, for example, disclose image pickup apparatuses capable of forming an image by obtaining image signals from different image pickup regions of a single image pickup element.
In Japanese Patent Laid-Open No. 09-214836, the image pickup apparatus has, as an operation of reading signals from the image pickup element, a skip mode in which pixels are thinned out and scanned, and a block mode in which a number of pixels (e.g., pixels in a predetermined region) are scanned but not thinned out. The image pickup apparatus performs scanning in each of the modes such that image data obtained from the scanning is read out on a frame-by-frame basis. Furthermore, in Japanese Patent Laid-Open No. 09-214836, images obtained through the skip mode and the block mode may be not only alternately read out on a frame-by-frame basis but also read out by changing a rate or frequency of a period of scanning in the skip mode to a period of scanning in the block mode.
In Japanese Patent Laid-Open No. 2000-032318, a method is disclosed for dividing a pixel array in which pixels are arranged in a matrix into a first portion and a second portion, reading signals from the first portion after pixels in the first portion are thinned out, and reading signals from all pixels in the second portion without thinning out the pixels in the second portion. In order to obtain an image in which the entire pixel array thereof is thinned out, among signals read from all the pixels, only signals of pixels in portions subjected to thinned-out scanning are read from a buffer memory arranged in a later stage of the image pickup element.
Japanese Patent Laid-Open No. 2007-150643 discloses a technique of realizing auto-focus processing that is performed faster than a period of a monitored image. As shown in FIG. 8B of Japanese Patent Laid-Open No. 2007-150643, when a plurality of auto-focus frames are provided, a sequence shown in FIG. 12 of Japanese Patent Laid-Open No. 2007-150643 is performed. Specifically, Japanese Patent Laid-Open No. 2007-150643 discloses a technique of repeatedly performing an operation of reading all of the plurality of auto-focus frames after a part of the monitored image is read.
For example, as shown in Japanese Patent Laid-Open No. 09-214836, in a case where low-density scanning is performed on pixels in an entire screen after thinning out and scanning are performed on all pixels in a specific region, the larger the number of pixels is in the image pickup element, the longer the time is required for reading the entire image. Therefore, as described in Japanese Patent Laid-Open No. 09-214836, when an image obtained in the skip mode and an image obtained in the block mode are alternately output, a speed of updating the image obtained in the block mode, which is a partial image that can be read at high speed, is limited by a speed of updating an entire image obtained in the skip mode, and thus the updating speed is lowered. When control of field angle and focus control are to be simultaneously performed, a high-quality image of a region of interest is required to be obtained and a refreshing speed is required to be high. According to the Japanese Patent Laid-Open No. 09-214836, the image of the region of interest may not be obtained at a sufficiently high refreshing speed.
Furthermore, when a relative rate or frequency of the skip mode to the block mode, which modes are to be switched from one to another, is changed, that is, when an operation of obtaining an entire image for a single screen after a partial image of interest is obtained for a plurality of screens (a plurality of frames) is repeatedly performed, an average refresh cycle of the partial image is reduced. However, because an update interval, i.e., a refresh cycle between two partial images obtained with the entire image being interposed therebetween, is different from refresh periods of partial images successively obtained without the entire image being interposed therebetween, there arises a problem in that movement of an obtained moving image has non-uniformities.
Moreover, in a case where a region to be read as the partial image is large and the entire image is considerably thinned out so as to be read with considerably low density, because the time required for reading the partial image is longer than the time required for reading the entire image, the refresh cycle of the entire image becomes large is limited by a speed of reading of the partial image. Under this condition, when the rate or frequency of the entire image to be read to the partial image to be read is changed, the problem described above arises. That is, there arises a problem in that a refreshing speed of an image that requires a shorter reading time is limited by a refreshing speed of an image that requires a longer reading time, and, accordingly, the refreshing speed of the image that requires the shorter reading time is deteriorated.
In the exemplary embodiment of Japanese Patent Laid-Open No. 2000-032318, because different regions in a certain single frame are subjected to different scanning operations, refreshing speeds of an entire image and a partial image are determined by a total reading time for the single frame. In this case, these refreshing speeds should be improved in order to make a refresh cycle of an image of interest shorter, which is difficult.
In addition, in an exemplary embodiment of Japanese Patent Laid-Open No. 2007-150643, assuming that a time period in which a part of the monitored image is read and a time period in which a plurality of auto-focus frames are read each corresponding to sub-frames, the following disadvantages arise. Specifically, when the monitored image is divided into four parts or portions, as shown in FIG. 12 of Japanese Patent Laid-Open No. 2007-150643, a period of eight sub-frames is required for updating the monitored image. On the other hand, a period of two sub-frames is required for updating the auto-focus frames. Therefore, a difference among exposure timings of the four portions obtained by dividing the monitored image becomes large. The larger the number of portions obtained by dividing the monitored image becomes, the larger the difference becomes.